Population density: not a primary factor in Covid-19

From a demo­graph­ic stand­point, what can we say about the Cov­id-19 crisis?

In my book Ser­ons-nous sub­mer­gés ?, which was pub­lished in Octo­ber 2020, I stud­ied the way the first wave of the pan­dem­ic unfol­ded, day after day, in four coun­tries – France, Switzer­land, Italy and Spain. I stud­ied them at the level of regions, counties and provinces. The res­ults showed that geo­graph­ic – rather than social – factors were key for dynam­ics of the epi­dem­ic. Cer­tainly, as many stud­ies have shown, immig­rants and low-income people have a high­er risk of dying from Cov­id-19 than the rest of the pop­u­la­tion. Not because of where they’re from or how much they earn, how­ever. Rather, it is because of their prox­im­ity to the vir­us. These pop­u­la­tions work in hos­pit­als, at super­mar­ket check­outs, as deliv­ery per­sons or taxi drivers. That is why they are more impacted. The vir­us can­not dis­tin­guish between an immig­rant and a non-immig­rant, nor can it tell how much money you make. Rather, it goes straight to the nearest per­son, and we can learn a lot more about who the nearest per­son will be through geography.

We often think of Cov­id-19 as an “urb­an vir­us”. How­ever, your study found that pop­u­la­tion dens­ity was not one of the key factors in its spread. Why is that?

When you com­pare Coronavir­us deaths between 1^st March and 15^th May 2020 with the dens­ity of French départe­ments, there are huge dif­fer­ences. Over that peri­od, the mor­tal­ity rate in Ter­ritoire de Belfort (1.18 per 1000 res­id­ents) was 170 times high­er than that of Ariège (.007 per 1,000). The con­trast between a map of mor­tal­ity rate and a map of pop­u­la­tion dens­ity shows no cor­rel­a­tion (see below). 

Dens­ity is not a factor in the large-scale spread of the vir­us. What mat­ters is the loc­a­tion of clusters, which are ini­tially linked to just one per­son. The more people are con­tam­in­ated before the infec­ted per­son real­ises, the harder it is to con­tain the epi­dem­ic. This is what happened with the second wave, but the dif­fer­ences in mor­tal­ity remain con­sid­er­ably dif­fer­ent – some­times as much as ten­fold high­er. And we see that “patient zer­os” appear in both the coun­tryside and the city includ­ing large towns such as Mul­house and Ajac­cio, but also smal­ler ones like Auray, Creil or even a vil­lage in Savoie, Les Contamines-Montjoie. 

Hence, while dens­ity mat­ters as the epi­dem­ic spreads, the ini­tial impact is min­im­al. The only thing that we can say is that patient zero is a trav­el­ler. So, they often appear near big inter­na­tion­al hubs (Geneva, Mil­an, Roissy Air­port near Par­is, New York, etc.). But they also travel out of the city, which is where the cluster devel­ops, i.e. Crépy-en-Valois, La Bastide-Mont­joie, Ber­ga­me. This is actu­ally what happened at the start of the AIDS epidemic. 

So to con­trol the spread one must first con­trol people’s move­ments – hence the lock­down. After expo­nen­tial growth of cases in the first clusters (Mul­house, Auray, Mil­an, etc.), the epi­dem­ic in the first wave was con­tained. For example, it prac­tic­ally didn’t get into the Loire at all, nor into Andalusia or South­ern Italy. It also explains why deaths were thirty times high­er in Mil­an than in Naples. It’s a remind­er of how, dur­ing the last out­break of the bubon­ic plague in France (Mar­seille, 1721), lines of sol­diers were deployed to pre­vent the epi­dem­ic from spread­ing bey­ond the Provence region.

Let’s go back to social cri­ter­ia. Which did you select for your study?

In each of the coun­tries included in our study, fig­ures for Cov­id-19 deaths were com­pared with four indic­at­ors: dens­ity, poverty, the pro­por­tion of immig­rants in the pop­u­la­tion, and the pro­por­tion of people over 70 years old. The geo­graph­ic dis­tri­bu­tion of these four factors gave us no indic­a­tion of which seg­ment of the pop­u­la­tion would be impacted. The maps speak volumes. This is due to the fact that the first wave was con­tained in the four coun­tries studied.

What about the second wave?

Para­dox­ic­ally, the second wave ori­gin­ated at the end of the first lock­down. The num­ber of daily cases was very low at the end of June. But people in the 15–49 age brack­et accoun­ted for two thirds of new cases. Many of them passed by undetec­ted, as they were asymp­to­mat­ic. With people trav­el­ling for the sum­mer hol­i­days, the vir­us spread all over France. Older people, par­tic­u­larly grand­par­ents, were also infected.

As such, in Octo­ber France found itself with a large num­ber of clusters. With the excep­tion of May­enne, these could not be con­tained as seen in Ajac­cio, Auray and Les Con­tam­ines-Mont­joie. As a res­ult, the epi­dem­ic spread pretty much every­where, as it had done in the two big clusters in the first wave in Creil and Mulhouse.

Because of more wide­spread safety meas­ures and improved health­care, the vir­us did not spread very quickly. The spread of Cov­id-19 across almost the entire coun­try has pro­duced new social dif­fer­ences con­nec­ted to what happened dur­ing the first wave – cer­tain groups are more care­ful than oth­ers, as shown by debates about mask-wear­ing. Nev­er­the­less, region­al dif­fer­ences are not insig­ni­fic­ant – between coastal Brit­tany and the Lyon region, the rate of cases and the mor­tal­ity rate var­ies by a factor of ten.

Just as we saw in the first wave, the second lock­down (pre­vent­ing people from trav­el­ling) will prob­ably main­tain these dif­fer­ences, once the wave has been stopped. Like oth­er con­tact epi­dem­ics through­out his­tory, con­trolling people’s mobil­ity remains an essen­tial factor for con­trolling the epi­dem­ic. We had bet­ter remem­ber that if we want to pre­vent a third wave.

Interview by Clément Boulle